Click here to return to Session Directory

dP

Dynamic
Positioning
Committee

Marine Technology Society

DYNAMIC POSITIONING CONFERENCE

October 12-13, 1999

SENSOR SESSION

GLONASS and GPS: Redundancy and Reliability

Obtained by Combining the Two Systems
James M. Naismith 1, MS, RPLS, LSLS
Gary A. Jeffress2, Ph.D., RPLS
Daniel Prouty2, MS
1. Javad Positioning Systems
San Jose, California
Ph. (281) 852-6036, (408) 453-2200
jim@javad.com
2. Conrad Blucher Institute for Surveying and Science
Texas A&M University - Corpus Christi, Texas
Ph. (361) 825-2720, Fax (361) 825-2795
jeffress@tamucc.edu, dprouty@cbi.tamucc.edu

Click here to return to Session Directory

Naismith, et al

Sensors

GLONASS and GPS

Abstract
The Global Positioning System (GPS) has been operational for several years. Even with a full satellite
constellation, the number of visible satellites at times drop to minimum levels for reliable positioning.
The Russian Global Navigation Satellite System (GLONASS) has also been operational for several years
and can provide the extra redundancy in satellite numbers when GPS coverage is at a minimum. The
GLONASS constellation was completed in 1996 and has been maintained at a level that makes it an ideal
supplement to the GPS.
The combination of tracking GLONASS satellites and GPS satellites provides sufficient data redundancy
to enable continuous reliable positioning. This paper describes the history, current status and future of
GLONASS. The datum PZ-90 is described along with its integration with the WGS84 datum. Postprocessing data from the GLONASS with data from the GPS is discussed.
A project in the Corpus Christi Bay area demonstrates results of using GLONASS and GPS in a marine
application. Coast Guard Beacon corrections, post-processing and real-time kinematic (RTK) positioning
methods using GPS/GLONASS receivers are discussed.
Given the recent surge of interest in using the GLONASS and the need for more redundancy in satellite
positioning systems, the combination of GPS and GLONASS can provide a viable solution.

DP Conference Houston

October 12-13, 1999

Page 1

Click here to return to Session Directory

Naismith, et al

Sensors

GLONASS and GPS

Introduction
Satellite based positioning has been in use for many years. During the cold war years the United States
and the Soviet Union developed satellite based navigation systems. These two systems are now being
used in combination to provide redundancy and increased reliability. When the two systems are used for
precise positioning instead of navigation, several factors need to be considered to eliminate systematic
error in the position.
Redundancy comes from the number of satellites and is easy to define. Reliability depends on many
factors, some of which are difficult to define. Reliability can be a matter of fact and measured or it can be
perceived reliability. Both are important in precise positioning applications. The reliability of the
position is heavily dependent on the reliability of the system. This paper provides a general background
of GLONASS and lists its deployment history to better define the perceived system reliability. Some
basic tests are reported to illustrate first, the ability to work with the GPS without causing reliability
problems and second, as a complement to the GPS to provide greater redundancy.

History of GLONASS
The History of GLONASS parallels the history of GPS. It began with the first launch on October 12,
1982. The satellites are launched three at a time on the Proton launch vehicle. A total of 82 satellites
have been deployed in 28 launches. The number of satellites deployed per year is shown in the following
chart.
10
9
8

SV's Deployed

7
6
5
4
3
2
1

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

1983

1982

Number of SVs deployed by year.

The first satellites launched were prototypes with an expected life span of two years. On average they
met or exceeded design life. The second set known as type IIc had a life expectancy of three years. The
average age of the last 12 satellites to be taken out of service was 44 months compared to the expected
life of 36 months. The current design is the GLONASS-M type with a life expectancy of five years.
GLONASS-M type satellites have not been launched yet so data on actual life span is not available. The
DP Conference Houston

October 12-13, 1999

Page 2

Click here to return to Session Directory

Naismith, et al

Sensors

GLONASS and GPS

following chart shows the total number of usable satellites since GLONASS came on line. A full
constellation of 24 satellites was achieved in 1996. Currently there are 15 usable satellites in the
constellation. (All data on numbers of launches, satellites and current status are taken from the KNITs
web-site).
30

Satellites

25
20
15
10

Jan-98

Jan-96

Jan-94

Jan-92

Jan-90

Jan-88

Jan-86

Jan-84

Jan-82

Number of usable satellites.

Current Status of GLONASS

As mentioned above there are currently 15 usable satellites in the GLONASS constellation. A full
constellation is considered 24 satellites. The last three satellites were put into service in early 1999. The
other twelve have been in use for over three years and would be expected to go off-line one at a time over
the next few years.
Data received from GLONASS is currently being used all over the world for many different types of
applications. GLONASS does not have selective availability, therefore autonomous positions are much
more accurate than those obtained from the GPS. The horizontal accuracy of autonomous positions from
GLONASS is <40m 95% according to ftp://ftp.nz.dlr.de/nav/glonass/status_inf/check_pos.gif for day
174,1999. This includes all positions. The actual horizontal accuracy can be much better if positions are
selected with low PDOP.

The Future
The future of GLONASS is a much-debated subject. There do not seem to be clear answers to the
questions that arise. This is due to the political questions that go unanswered in the countries that made
up the former Soviet Union. These political decisions have a direct impact on the funding and direction
of the GLONASS. Although launches have been scheduled only to be delayed by lack of funding, the
system has been maintained at above ten satellites since 1991.
The Ministry of Defense of the Russian Federation is responsible for the operation of GLONASS. The
Russian Space Agency is responsible for civil user application and development. In February and March
of 1999, the Russian Federation expressed its willingness to cooperate with the international community
and its willingness to use the GLONASS as a basis for development of an international satellite

DP Conference Houston

October 12-13, 1999

Page 3

Click here to return to Session Directory

Naismith, et al

Sensors

GLONASS and GPS

navigation system. The federation also expressed its commitment to the funding of GLONASS and to
bring foreign investment in the system (KNITs).
We do not know what the future holds for the Russian economy. Even with the current lack of funding,
the willingness of the Russian Federation to move forward does not seem to be hampered. They have
proven their ability to work with what is available, and show no signs of giving up. Given the past
history and current status of GLONASS, it does appear that the Russian Federation is committed to
maintaining it.

PZ-90
Parametry Zemli 1990, Parameters of the Earth 1990, also known as PE-90, is the reference frame and
ellipsoid used by GLONASS. PZ-90 was formerly the Soviet Geodetic System 1985/1990
(SGS85)(DLR). The semimajor axis of the PZ-90 ellipsoid is 6378136m and flattening (1/f) is
298.257839303. It is used in the same way as WGS84 is for the GPS. The semimajor axis of WGS84 is
6378137.0m and flattening (1/f) is 298.257223563.
SGS85 was in close alignment with WGS84. A rotation around the Z-axis of 0.6 seconds and a
displacement in the Z direction of 4 meters brought the two systems into coincidence (Misra, et al, and
SGS85). PZ-90 is also in close alignment with WGS84. A rotation around the Z-axis of 0.4 seconds
and a displacement along the Y-axis of 2.5 meters brings the two systems into coincidence (Misra, et al,
and Integrated). Beginning in early 1999, additional data became available from the IGEX
(International GLONASS Experiment) campaign.

Post-processing and Real-time processing with GPS and GLONASS

Processing GLONASS data is very similar to processing GPS data. Like GPS, GLONASS is a dual
frequency system with C/A code on L1 and P-Codes on carriers L1 and L2. Unlike GPS, GLONASS
uses frequency division multiple access (FDMA). Each of the GLONASS satellites transmits the same
code on different frequencies. GPS uses code division multiple access (CDMA). Each of the GPS
satellites transmits a different code on the same frequency (Leick).
There are different time delays in receiver hardware for the different frequencies used by the GLONASS
vs. GPS. The instantaneous time at which GPS and GLONASS measurements come in are slightly
different. These differences are estimated for GPS/L1 - GPS/L2, GPS/L1 - GLONASS/L1 and GPS/L1 GLONASS/L2. The error due to the delays is not very large and is almost continuous, depending on the
individual receiver. It can be estimated using a conservative least squares estimator and its effects taken
into account (Rapaport).
For RTK applications the additional satellites observed when using both systems are useful for faster,
more robust integer ambiguity resolution (Leick). The amount of computations needed to solve the
integers grows exponentially with the number of satellites. For this reason it is useful to solve for a
subset of the total number of integers and then resolve the others (Rapaport).

DP Conference Houston

October 12-13, 1999

Page 4

Click here to return to Session Directory

Naismith, et al

Sensors

GLONASS and GPS

Application
The first test was run in the Port of Corpus Christi, Texas. A hydrographic survey of a dock was used.
The objective of the test was to determine if there would be any difficulties using a GPS/GLONASS
receiver with Differential GPS only RTCM corrections from a Coast Guard Beacon. A Trimble
DSM212L receiver and a Javad Legacy receiver were used. Both systems accepted Coast Guard Beacon
corrections from the station at Port Aransas 35km away. The test was run with the JPS receiver collecting
dual frequency GPS and GLONASS data. Dual frequency, dual system data for post-processing were
simultaneously collected at a local benchmark. Differential corrections for GLONASS were not available
for this test from the same source. Positions are compared between the JPS receiver corrected with data
from the beacon and the post-processed data. The positions obtained by post-processing were taken as
"true" for purposes of comparison. The Trimble data was used as a hardware independent check. All
positions were reduced to NAD83 State Plane Coordinates, Texas Southern Zone.
The survey route is shown in the following figure. The post-processed data is represented by red triangles,
JPS positions by green circles and Trimble positions by yellow squares.

Survey Route.

DP Conference Houston

October 12-13, 1999

Page 5

Click here to return to Session Directory

Naismith, et al

Sensors

GLONASS and GPS

Relative Positions.
The differences between the positions appear to be systematic and smooth. Over 700 positions were used
to calculate the average differences between the positions obtained by differential and post-processing.
North

East

North

East

JPS - PP

JPS - PP

JPS - TRI

JPS - TRI

Average (m)

1.04

-0.02

0.70

-0.53

Stdev (m)

0.86

0.61

-1.06

-0.64

JPS = Javad Positioning Systems Legacy/LegAnt

PP = Post-Processed
TRI = Trimble DSM212L
The results were as expected with standard deviations of less than one meter. Errors appeared to be
systematic and changed when a satellite was added or lost in the solution.
Both receivers received corrections for GPS satellites. If a satellite is tracked at the rover and it is not
tracked at the base, no corrections are available for it. The receiver simply does not use the data from that
satellite. Since the GLONASS satellites did not have corrections for them, they were simply ignored as

DP Conference Houston

October 12-13, 1999

Page 6

Click here to return to Session Directory

Naismith, et al

Sensors

GLONASS and GPS

though they were GPS satellites with no corrections. The GLONASS satellite's data were used in the
post-processing.
A second test was performed on a static point in an area with significant multipath and obstructions. A
single antenna was connected to two receivers, one using GPS only and the other using GPS and
GLONASS. A base receiver was set up nearby and transmitted the corrections via RTCM format. Data
was collected for 24 hours to provide a sample with varying satellite constellation and night vs. day.
Preliminary results show that 98.5% of the dual system data had a standard deviation of one meter. 96.5%
of the GPS only data had a standard deviation of one meter. There were 0.5% fewer positions generated
with GPS only. The following figure is a plot of the data. The blue circles are GPS only and the black
plus signs are dual system. The red circle in the center is the position taken as true. It was estimated by a
post-processed static baseline.
All positions were plotted up to a PDOP (Positional Dilution of Precision) of 30. PDOP provides a good
measure of the geometry of the satellites in relation to the point being positioned. Values greater than 7
or 8 indicate geometry that is not good for reliable positions. Areas with sky blockage such as was used
for this test can have severe problems with satellite geometry because only parts of the sky are visible.

Plot of Positions.

DP Conference Houston

October 12-13, 1999

Page 7

Click here to return to Session Directory

Naismith, et al

Sensors

GLONASS and GPS

The next figure shows the same data at a much larger scale. The difference between GPS only and dual
system is more apparent. Setting a maximum PDOP could have filtered out the positions that are far from
the true point. This would have created gaps in the data that are not present using both systems.

Plot Showing GPS Positions with High PDOP.

The following are plots of PDOP vs. time for the two datasets. Using GPS only there are several spikes
in PDOP where the position is unreliable, some up to one hour in length. Using both systems, the spikes
are much fewer and very short in duration.

DP Conference Houston

October 12-13, 1999

Page 8

Click here to return to Session Directory

Naismith, et al

Sensors

GLONASS and GPS

GPS only.

GPS and GLONASS.

DP Conference Houston

October 12-13, 1999

Page 9

Click here to return to Session Directory

Naismith, et al

Sensors

GLONASS and GPS

Conclusions
There have been many articles on using the GLONASS as a standalone or as a compliment to the GPS.
With the continuous changes in both systems it is worth revisiting the data and evaluating the reliability
of the systems on a periodic basis. Using GLONASS in conjunction with GPS does not appear to cause
any problems or errors when corrections are available for GPS only. Using GLONASS in conjunction
with GPS provides more reliable positioning. The long-term future of GLONASS is dependent on many
factors that have little to do with satellite navigation. The short-term future of GLONASS appears stable
when using the past history as a guide.

DP Conference Houston

October 12-13, 1999

Page 10

Click here to return to Session Directory

Naismith, et al

Sensors

GLONASS and GPS

References
Ashjaee, Javad, Javad Positioning Systems, Personal interview, August, 1999.
KNITs: Coordination Scientific Information Center web site at
http://mx.iki.rssi.ru/SFCSIC/csic.html
DLR Neustrelitz Remote Sensing Ground Station web site at http//:www.nz.dlr.de/gps (German
Aerospace Center)
Leick, Alfred. GPS Satellite Surveying, Second Edition. John Wiley & Sons, Inc. New York.
1995.
Misra, M. Pratt, R. Muchnik, B. Burke, and T. Hall, GLONASS Performance: Measurement
Data Quality and System Upkeep, http://vega.atc.ll.mit.edu/papers/GLO-96/glonass96.html, 1996.
MIT Lincoln Laboratory GLONASS Group web site at http://vega.atc.ll.mit.edu (no longer
maintained).
Misra, P., M. Pratt, and R. Muchnik. GLONASS Performance in 1994: A REVIEW,
http://vega.atc.ll.mit.edu/papers/GLO-94/glonass-94.html .
Misra, P., M. Pratt, and R. Muchnik. GLONASS Performance in 1995: A REVIEW,
http://vega.atc.ll.mit.edu/papers/GLO-94/glonass-95.html .
Misra, P., R.I. Abbot and E.Bayliss, SGS85 - WGS84 Transformation
http://vega.atc.ll.mit.edu/papers/PZ90-WGS84/PZ90-WGS84.html .

Misra, P., R.I. Abbot and E.M. Gaposchkin, Integrated use of GPS and GLONASS:
Transformation Between WGS 84 and PZ-90, http://vega.atc.ll.mit.edu/papers/PZ90WGS84/PZ90-WGS84.html .
Rapaport, Lev, Javad Positioning Systems, Moscow, Personal Interview via e-mail, August,
1999.
Slater, James A. et al, The International GLONASS Experiment (IGEX-1998) Proceedings of
ION GPS-98, Sept. 15-18, 1998.
Zinoviev, Alexei, Javad Positioning Systems, Moscow, Personal interview via e-mail, August,
1999.

Version August 30, 1999

DP Conference Houston

October 12-13, 1999

Page 11